Thursday, June 4, 2009

Broadband Stimulus

We need good, accessible, inexpensive broadband

Shyam Ponappa / New Delhi June 4, 2009

Does India need a broadband stimulus plan? Yes, because as in other countries, it can be a force-multiplier throughout. Take education: broadband can make centrally deployed content and teaching skills accessible at all levels — schools, undergraduate and postgraduate education, vocational studies, and Continuing Education/adult education. It can help bring about the inclusive growth the government talks of. It facilitates information use by making aggregation, coordination, access/dissemination and management of delivery more effective and efficient.

The potential public benefits extend to healthcare and governance, starting from law-and-order and security, extending to energy management, water and sanitation, tax collection and subsidies, commerce, industry and services, including information, entertainment, logistics, etc. Extensive benefits can be obtained from limited investments. However, policies and regulations can make or break the industry, or even the political economy, because it can take 10 to 20 years to get returns from investing in fibre and wireless network systems. All those enamoured with lower tariffs should apply their minds to (a) what we need, (b) for what purposes, eg, the gamut of education, healthcare, other public services… (c) with the binding constraint of sustainable returns.

A Low-Cost Stimulus!

But here is the amazing thing: if appropriate policies and incentives are devised, as we have seen with India’s telecommunications since the New Telecom Policy, 1999 (NTP ’99), private and state-owned enterprises can build the networks and provide services. The government need not invest directly. Instead, it must provide clarity of objectives, enabling policies and procedures, and institutional support (kept fair and incorruptible), as well as foregone up-front revenues because of tax breaks and other incentives. As with telecom, this can result in explosive growth and increased government revenues.

To see this in perspective, compare America’s investment in its interstate highway system with its broadband stimulus package. The highway system was built over generations; one estimate in 2006 puts it at 35 years and $425 billion. President Obama’s economic stimulus package of $819 billion has a broadband component of just $7.2 billion over two years (under 1 per cent). For a start, America’s Federal Communications Commission (FCC) is required to submit a plan to Congress by February 2010.

Of course, India is very far from having America’s systems and therefore needs to do much more. However, the point is that the US government plans little direct investment in broadband. Second, even with its can-do culture, America is allowing a year for just developing recommendations. Moral: India cannot expect a magical fix in weeks.

Reality and International Benchmarks

Many governments are striving to stimulate growth and productivity through broadband. A study by Oxford’s Saïd Business School and the University of Oviedo last year covered users in 42 countries. Figure 1 shows the ‘Broadband Quality Score’, a composite of user download speeds, upload speeds and latency.

Figure 1: Broadband Quality Scores by Country


India was the tail-ender with a BQS close to 0 on a scale of 100 (the international average is 32), despite our IT skills and knowledge economy aspirations. Another takeaway is that to compete, we need initiatives to prepare for a much higher level of service provision and usage (ie, for an average future BQS of 75 compared with the current average of 32).

How They Did It

In countries like Japan, Sweden and Korea, broadband development resulted from a combination of government policies and incentives, including subsidies for building infrastructure as well as for demand and use, eg, paying schools and hospitals so they would use high-speed services.

Japan used accelerated depreciation, low-cost loans and subsidies for rural networks. Sweden provided tax breaks, rural subsidies, and required state-owned utilities to create backbone networks, reducing costs for local service providers.** Korea and other southeast Asian countries not in this study — Hong Kong, Taiwan, Singapore, Malaysia — have similar government policy support for broadband.

Objectives: Good, Accessible, Inexpensive Services

Our starting point should be clear objectives for broadband services in a realistic time frame. The cornucopia of potential public benefits is enticing. But to reap these benefits, broadband must be (a) ubiquitous, ie, accessible everywhere — or in most places, for a start — and (b) be reasonably priced.

Once these objectives are established, there remain the tasks of planning and execution. The government will need to formulate convergent policies that are coordinated across ministries and states. If it adopts a systematic process of backward induction, developing sound, disaggregated task plans with specific resources, time lines, and responsibilities for delivery and — this is especially important — good quality standards, the objectives may be achieved.

The government will need to avoid getting caught again in specifying technologies, learning from the cellular experience, while focusing on service standards and pricing targets. It must also accept that because of network economics, communications networks cannot support unrestricted competition with forced access. Rather, we must have a restricted number of players, and optimise services and pricing.

Implications for Action

What does India need to do going forward?

  • Define appropriate objectives, eg, broadband (at least 2 Mbps) everywhere at reasonable prices.
  • Design incentives, eg, a reduction in revenue-share on cellular revenues for incremental broadband service usage (not just service provision) at stipulated levels and locations (geographies). If the government accepts this approach in principle, a consultative process with experts and stakeholders could evolve reasonable incentives that are workable in the public interest.
  • Accept the economics of networks and strive for limited competition with good regulation.
  • Avoid imposing arbitrary costs such as auctions and up-front fees, whether for spectrum or for network franchises. Go with proven revenue-sharing to achieve service delivery.
  • Stay technology-agnostic and specify service levels including quality, and strive for objective regulation for reasonable pricing. For instance, J:Com, a cable service in Japan, apparently can provide 160 Mbps connections in its service area for an incremental cost of $20, possibly the lowest cost worldwide. Similarly, issues such as shared spectrum use, especially of lower frequencies (under 1 GHz), need expert consideration and evaluation. 

* Also see: ‘Broadband Quality Matters!’, Benoit Felten:  

** For details, see Saul Hansell: